# valiant-mtl
MTL-style adapter for
[`valiant`](https://hackage.haskell.org/package/valiant).
All operations work in any monad with `HasPool m` (a `MonadReader`-like
class that exposes a `Pool`) and `MonadIO m`. Use this when your
application lives in an `mtl` stack with a richer environment than
just `Pool`.
## Quick start
```haskell
import Control.Monad.Reader
import Valiant (newPool, defaultPoolConfig, poolConnString)
import Valiant.Mtl
data AppEnv = AppEnv
{ appPool :: Pool
, appLogger :: Logger
}
instance HasPool (ReaderT AppEnv IO) where
getPool = asks appPool
myApp :: (HasPool m, MonadIO m) => m [User]
myApp = do
users <- fetchAllMtl listUsers ()
count <- fetchScalarMtl countUsers ()
pure users
main :: IO ()
main = do
pool <- newPool defaultPoolConfig { poolConnString = "postgres://..." }
logger <- newLogger
runReaderT myApp (AppEnv pool logger)
```
## What you get
- `class HasPool m where getPool :: m Pool`
- Query operations: `fetchOneMtl`, `fetchAllMtl`, `fetchScalarMtl`,
`fetchOneOrThrowMtl`, `fetchExistsMtl`, `forEachMtl`
- Command operations: `executeMtl`, `executeReturningMtl`,
`executeBatchMtl`
- Transactions: `withTransactionMtl`, `withTransactionLevelMtl`
- Pool helpers: `poolStatsMtl`, `withConnectionMtl`
There is also a `Valiant.Monad` module re-exported from this package
that provides the same operations specialised to
`ReaderT Pool IO`.
See the [valiant tutorial](https://github.com/joshburgess/valiant/blob/main/docs/TUTORIAL.md)
for `Statement` definitions and the `valiant prepare` workflow.